I am going to second the idea of buying a kit.
Dan
Kits containing detailed information, such as those that Al the Fisher Doc offers (other issues aside), are very good for the novice. With those kits, you are paying for the detailed information and guidance and not so much the parts.
Kits containing just parts, or perhaps with a basic (now virtually public domain) schematic thrown in (like the old VTV kits), are not very good values IMO, except for the very lazy.
The typical reconditioning cost, even using high quality parts like Atoms, Dales, and CDEs, if designer caps are not used, is still around $25. Most kits do not offer better parts quality, and in many instances are worse, than what can be bought from Mouser for a small fraction of the kit cost.
Kits may also not comport well with other service options, such as the SDS cap board. You may find some kit parts useless (and wasted) in those situations.
You're also at the mercy of what the kit designer thinks needs replacing, and at what values. And you're also at the mercy of the kit designer's competence. I've seen supposedly reputable kits specifying totally inappropriate (and unsafe) DC film capacitors as suppressor replacements, where a Y-Class part is required.
There is a lot of information on these instruments all over the Internet at this point, on exactly what needs checking, repairing and replacing. Plenty of options are out there to consider for someone willing to look it up. An hour's worth of research is usually worth the money saved on a kit, and can result in a far more satisfying project result.
At one point about six or eight years ago, I started drafting a comprehensive FAQ/restoration guide for the Fisher tube receiver family, including detailed parts lists and parts brand listening tests and recommendations. But I never finished it. Too much else to do. I guess if I ever did get around to finishing it, it could be a publicly posted "project" on Mouser. One click, and your parts are ordered. But I guess all the Internet kit sellers would have a fit if that happened. Nevertheless, there's still enough specific data posted out there to quickly cobble together exactly what is needed.